Agent-based modelling is the generic designation of a family of computational techniques used in the simulation of the behaviour of a number of agents acting as decision-makers in a given environment. The agents are the key elements of the theory; they represent those actors that, during simulation, interact in a given environment. The behaviour of such agents is usually typified in terms of learning/adaptive rules and simple behaviours (mostly reactive); their interactions “create” complex dynamics along hinged structures.
This is often referred to a self-organization process: given the learning and behavioural rules, the agents organise spontaneously and autonomously, without external intervention, nor any central structure (Niglia et al., 2012). Moreover, the advancement of computational techniques pushed these models to an even higher level of complexity. Regarding the theoretical foundation, the most interesting and widespread agent architecture is the Belief-Desire-Intention (BDI) architecture, introduced by Bratman as a philosophical model for describing rational agents (Bratman 1987). It consists of the concept of belief, desire and intention as mental attitudes, to mimic human actions. Beliefs capture informational attitudes, desires motivational attitudes, and intentions deliberative attitudes of agents. Rao and Georgeff (1995) have adopted this conceptualisation, built their formal theory and an execution model for software agents based on the notion of beliefs, goals, and plans.
The typical developed model is based on JADEX technology employed to facilitates the use of BDI models in the context of mainstream programming. Introducing beliefs, goals and plans as first class objects enables the manipulation “inside” the agent and involves non-predictable dynamics without top-down planning and therefore, increases the analytical power of the model when an emergent phenomenon is unexpected and not-predictable. It can be an effective tool to study complex systems and processes of emergence and self-organization for policy evaluation.